Background
Angiogenesis, the formation of new blood vessels, is an essential physiological process in embryo development, normal growth, and tissue repair, and is tightly regulated at the molecular level. Dysregulation of angiogenesis occurs in various pathological conditions and is one of the hallmarks of cancer[
1,
2]. Recognition of the role of angiogenesis during neoplastic development is important for a more comprehensive understanding of the mechanisms involved in tumor growth and metastasis[
3‐
5].
Tumor vessels are histopathologically different from normal vessels; most tumor vessels have irregular diameters, abnormal branching patterns, and do not fit well into the usual categorization of arterioles, capillaries, or venules[
6‐
8]. Moreover, the endothelial cells (ECs) that make up tumor vessels are often loosely interconnected and have intercellular openings and abnormal pericytes, contributing to the leakiness of these vessels[
9‐
11]. Structural abnormalities in the basement membrane of tumor vessels are also responsible for their relative immaturity in comparison with normal vessels[
12,
13].
Although the main purpose of tumor angiogenesis is to maintain blood supply to the tumor, the process usually occurs in an abnormally regulated fashion and the resulting tumor vasculature may have abnormal organization, structure, and function[
1]. Recent advances have led to a better understanding of the vascular changes in malignant tumors, but the structural abnormalities in blood vessels of benign tumors (pre-neoplastic lesions) and those present during malignant transformation are still poorly understood.
Angiogenesis is regulated by highly coordinated functions of various proteins that play pro- or anti-angiogenic roles[
14]. Pro-angiogenic factors include vascular endothelial growth factor (VEGF), fibroblast growth factor, platelet-derived growth factor, insulin-like growth factor, transforming growth factors, angiopoietins, and several chemokines, while anti-angiogenic factors include thrombospondin-1, angiostatin, and endostatin[
15]. Two novel endogenous paracrine factors, termed vasohibins, have also been described recently[
16]. Vasohibin-1 is anti-angiogenic, while vasohibin-2 appears to be pro-angiogenic. Vasohibin-2 (VASH2) is mainly expressed by infiltrating bone marrow-derived mononuclear cells at the angiogenic-sprouting front[
16‐
20], and the expression of VASH2 in human serous ovarian adenocarcinoma and hepatocellular carcinoma accelerates tumor growth by promoting angiogenesis[
21,
22]. However, it is not known if there are differences in VASH2 expression between tumor cells and tumor-associated ECs. In addition, although VEGF-targeted therapy shows promise for the inhibition of angiogenesis during tumor progression, new therapeutic targets are needed to advance anti-angiogenic treatments in cancer. Therefore, the development of novel therapeutic agents may be facilitated by identification and characterization of new angiogenic factors, and this may be achieved using spontaneous tumor models[
22].
Elucidation of the angiogenic patterns in benign tumors and the involvement of various pro- and anti-angiogenic factors are important for understanding the development of malignant tumors and the neoplastic transformation sequence in intestinal epithelia[
23,
24]. In the current study, relationships between tumor angiogenesis and multi-step carcinogenesis were assessed using the
Apc
Min/+
mouse model, which spontaneously develops multiple intestinal adenomas that mimic those that undergo early transformation into adenocarcinomas in patients with familial adenomatous polyposis[
25]. Using this mouse model offers the advantage of close recapitulation of the histopathological characteristics observed in human cancer. Furthermore, tissue-specific induction of mutations gives rise to orthotopic tumors in the context of a functional, immune-competent microenvironment, and thus includes the crosstalk between an emerging tumor and its environment[
26‐
28].
In the present study, we sought to characterize the microvascular changes that occur during the adenoma-carcinoma sequence in a tumor to determine whether pre-existing vascular changes can be used to predict tumor transformation from benign to malignant. For comparison, wild-type (WT) C57BL/6 mice and Vash2
-/-
mice bearing transplanted syngeneic CMT93 colorectal carcinoma cells were included in the study. The effects of VASH2 on adenoma growth and progression to carcinomas were also examined by cross-breeding mice with a complete absence of Vash2 expression (Vash2
-/-
mice) with Apc
Min/+
mice. Our results indicate a novel role for VASH2 in tumor angiogenesis as an index of malignant transformation and suggest that VASH2 may be a novel target for anti-angiogenic agents in cancer therapy.
Discussion
In this study, we sought to characterize microvascular changes in the intestine of
Apc
Min/+
mice, a useful animal model for studying spontaneous adenomatous polyposis and subsequent adenocarcinoma (a process termed as the adenoma-carcinoma sequence)[
24,
25]. We found that changes in local vascular networks reflected the neoplastic transformation sequence of the intestinal epithelia, as follows: First, both structural and functional changes in local vascular networks had already been initiated in benign tumors in
Apc
Min/+
mice, and in corresponding human surgical specimens. Second, the pattern of tumor angiogenesis in benign tumors was similar to that in malignant tumors. Third, VASH2 expressed in the adenocarcinoma cells promoted tumor growth and tumor angiogenesis in
Apc
Min/+
mice. Based on these data, we propose that the sequence of phases during tumor transformation from benign to malignant is based on microvascular changes, as summarized in Table
1.
Table 1
Changes in tumor microvasculature during multistep carcinogenesis
Morphological abnormality; | | | | |
Vessel density | - | 1+ | 2+ | 3+ |
Lumen | Round | Not round | Not round | Not round |
| Systematic | Tortuous | Tortuous | Tortuous |
Branching | - | 1+ | 2+ | 3+ |
Irregularity | - | 2+ | 3+ | 3+ |
Ec | | | | |
Morphology | Normal | Protuberance | Protuberance | Protuberance |
| | Microvilli | Microvilli | Microvilli |
IHC Lectin | + | + | + | + |
CD31 | + | + | + | + |
CD105 | - | ± | + | + |
BM | | | | |
Morphology | Single | Multiple | Multiple | Multiple |
Pc | | | | |
Morphology | Attached | Detached | Detached | Detached |
| | Protuberance | Protuberance | Protuberance |
IHC Desmin | + | + | + | + |
α-SMA | - | - | ± | + |
Pv | | | | |
IHC VASH2 | - | - | + | + |
Functional abnormality; | | | | |
Permeability | - | - | + | ++ |
Blood flow (lectin) | Normal | Irregularity 1+ | Irregularity 2+ | Irregularity 2+ |
Hypoxia (HIF1-α) | - | - | ± | + |
Gene mutation |
Apc
|
Apc
|
K-ras
| p53 and more |
Stage of microvasculature | Stage 0 | Stage I | Stage IIa | Stage IIb |
Morphological changes in blood vessels occurred earlier than malignant changes in the epithelium of intestinal lesions. This suggests that angiogenesis patterns may play a critical role in the development and growth of benign tumors during multi-step carcinogenesis. There is increasing evidence that angiogenesis may also play a critical role in development of benign tumors[
29‐
32]. Although angiogenesis is tightly regulated at the molecular level, dysregulation of angiogenesis is a hallmark of cancer and can lead to various pathological conditions[
33]. The imbalance of pro- and anti-angiogenic signaling within tumors creates an abnormal vascular network that is characterized by dilated, tortuous, and hyperpermeable vessels[
34]. The physiological consequences of these vascular abnormalities include temporal and spatial heterogeneity in tumor blood flow and oxygenation and increased interstitial fluid pressure in tumors[
35,
36]. These abnormalities and the resulting microenvironment fuel tumor progression and also lead to a reduction in the efficacy of chemotherapy, radiotherapy, and immunotherapies[
37]. However, none of these previous studies have demonstrated that both structural and functional changes in blood vessels during multi-step carcinogenesis reflect the neoplastic transformation sequence of epithelia and the degree of malignancy, as we observed in this study. We were able to show changes in tumor vessels during multistep carcinogenesis in spontaneous tumors as a whole vascular network using fluorescent 3D imaging and transmission electron microscopy. These results showed that the histopathology of the vasculature in late stage adenoma was similar to that of malignant tumors.
VASH1 is expressed in ECs in the termination zone, suppressing angiogenesis, whereas VASH2 is expressed mainly in infiltrating bone marrow-derived mononuclear cells at the sprouting front, promoting angiogenesis[
16,
19]. However, exogenous VASH2 exhibits anti-angiogenic activity in the mouse cornea[
17]. VASH2 expression has been demonstrated in certain ovarian cancers, where it promotes tumor growth and peritoneal dissemination of tumor cells by stimulating tumor angiogenesis[
21]. VASH2 is also highly expressed in hepatocellular carcinoma cells (HCCs) and tissues, and promotes HCC angiogenesis and malignant transformation[
22]. In our study, VASH2 was mainly expressed by late stage adenoma and spontaneous adenocarcinoma cells around tumor vessels in
Apc
Min/+
mice. Transplanted CMT93 tumors in
Vash2-/- mice were less vascularized and more regular than those in WT mice. Furthermore, in gastrointestinal tumors of
Apc
Min/+
/
Vash2
-/-
mice, the number of small intestinal polyps was significantly reduced, pericyte coverage of tumor vessels was increased, and tumor lesions were less vascularized than hyperplasia lesions. These results support the hypothesis that VASH2 plays an important role in tumor angiogenesis and tumor progression. Because inhibition of VASH2 normalized abnormal tumor vessels in adenocarcinoma, VASH2 may be an important therapeutic target in the treatment of human cancers.
Control of angiogenic factors such as VASH2 at stages of benign tumorigenesis may inhibit malignant transformation of the epithelium. If new anti-vascular agents such as anti-VASH2 neutralizing antibodies could be developed to suppress changes in local vascular networks, the intestinal epithelium may not become malignant during the adenoma-carcinoma sequence. Thus, the findings of this study may contribute to the development of new antivascular agents as prophylactic medicines for malignant cancers. In addition, we propose that the microcirculation may act as an index of malignant transformation and may have potential use in future diagnosis and treatment of cancer. A better understanding of the various mechanisms of angiogenesis will facilitate development of novel anti-vascular therapies for the treatment of malignant tumors.
Methods
Mice
C57BL/6 J-
Apc
Min/+
(
Apc
Min/+
) of both sexes were purchased from the Jackson Laboratory (Bar Harbor, ME) and male C57BL/6 (wild type: WT, MHC class I type: H-2
b) were purchased from Japan SLC Inc. (Shizuoka, Japan).
Vash2 knockout (
Vash2
-/-
) mice were generated as described elsewhere[
19].
Apc
Min/+
mice of a pure C57BL/6 background were mated to
Vash2
-/-
mice of a mixed C57BL/6 background, and the resulting pups were screened for the Min mutation[
38] and for the
Vash2
-/-
gene by PCR[
19,
39]. Mice were maintained in air-filtered clean rooms and fed sterilized standard laboratory chow and water
ad libitum. Because it has been previously shown that a combination of a high-fat diet and dextran sodium sulfate strongly promotes intestinal carcinogenesis in
Apc
Min/+
mice[
24,
25,
40], the animals were fed with a high-fat diet (Oriental Yeast, Tokyo, Japan) and 2% DSS (Tokyo Chemical Industry, Tokyo, Japan) to ensure development of adenomas and adenocarcinomas in the small intestine 3 months later. The Animal Experiment Committee, Tokyo Women’s Medical University (TWMU) approved the procedures employed in the handling and study of the mice. The following experiments were performed in accordance with legislation of the Institute of Laboratory Animals for Animal Experimentation at TWMU. Unless otherwise stated, at least 20 mice in each experimental group were examined.
Cell culture
CMT93 cells[
41] derived from a mouse rectal carcinoma were obtained from the European Collection of Cell Cultures (Sigma-Aldrich, St Louis, MO, USA). CMT93 cells were grown in Dulbecco’s modified Eagle medium (DMEM: Invitrogen, Carlsbad, CA, USA) supplemented with 10% heat-inactivated fetal bovine serum, 5 × 10
-5 M 2-mercaptoethanol, 10 mM HEPES, 1 mM sodium pyruvate, 3.75 g/L NaHCO
3, 2 mM glutamine, 100 U/mL penicillin, and 0.1 mg/mL streptomycin (Gibco, Grand Island, NY, USA). CMT93 tumor cells (2.5 × 10
6 cells) in 250 μL of calcium- and magnesium-free phosphate-buffered saline (Ca
++-, Mg
++-free PBS; pH 7.4) were injected into the dorsal subcutis (s.c.) of WT C57BL/6 mice and
Vash2
-/-
mice.
Definition of adenomas and adenocarcinomas
All tumors (early-stage: approximately 12 weeks or earlier, late-stage: approximately 16 weeks or later) that developed in mice were examined in paraffin sections stained with H&E, and histopathological changes such as carcinoma in situ and stromal invasion were evaluated. Tumors were diagnosed as adenomas by expansion to the mucosal layer, reduction of goblet cell numbers, and moderate loss of mucosal architecture by glandular growth and dilated cysts. Adenomas with 50% of high-grade dysplasia (severe distortion of the glandular architecture and prominent atypical cells) were considered carcinomas in situ. However, only the lesions showing invasion through the lamina muscularis mucosae were identified as adenocarcinomas.
General tissue preparation
All mice were anesthetized by an intramuscular (i.m.) injection of ketamine (87 mg/kg) and xylazine (13 mg/kg). Under deep anesthesia, the chest was opened and the aorta was perfused with 4% paraformaldehyde (PFA) in PBS for 10 min at a pressure of 120 mm Hg using an 18-gauge cannula inserted via the left ventricle. The blood and fixative were then flushed out through an opening in the right atrium. After perfusion, tissues were removed, cut into small pieces and rinsed in PBS, then further immersed in PBS containing a graded series of sucrose (up to 30%) at 4°C overnight. Subsequently, these tissues were embedded in Tissue-Tek O.C.T. compound (Sakura Finetek, Torrance, CA, USA) and snap-frozen in liquid nitrogen. Cryostat sections (14–120 μm) were cut, placed on silane-coated glass slides, air-dried for at least 2 h and then immunostained.
To obtain semi-thin Epon-embedded sections, various tissues were excised, cut into small blocks, and fixed by immersion in 2% glutaraldehyde in 0.1 M phosphate buffer (PB; pH 7.2) at 4°C for 24 h. After washing out the fixatives with 0.1 M PB, the blocks were treated with 1% osmium tetroxide (OSO4)-0.1 M PB (a mixture of 2% OsO4 + 0.2 M PB). The tissues were dehydrated in a graded series of ethanol, infiltrated with propylene oxide, and embedded in Epon. Semi-thin sections (0.5 μm thick) were made and stained with 1% toluidine blue in PBS.
Labeling of blood vessels with tomato lectin for 3d imaging
To identify blood vessels, we used the intravascular perfusion of fluorescent tomato lectin to label all blood-circulating vessels[
42]. Briefly, under anesthesia, the mice were intravenously (i.v.) injected with 100 μl of FITC-conjugated tomato lectin (
Lycopersicon esculentum lectin; 1 mg/mL; Vector Labs, Burlingame, CA). Tomato lectin binds uniformly to the luminal surface of ECs[
43] and can be used to label all blood vessels that have a patent blood supply. After perfusion, the tissues were processed for subsequent analyses as described above.
Immunohistochemistry
Cryosections were first incubated in 4% Block Ace (Dainippon Seiyaku, Osaka, Japan) to block nonspecific background stains, and successively incubated with various primary antibodies (alone or in combination) in PBS containing 1% bovine serum albumin (Sigma-Aldrich, St Louis, MO, USA) at 4°C overnight. ECs were identified with antibodies to CD31 (PECAM-1; hamster monoclonal, 1:400; Chemicon, Billerica, MA, USA) and Von Willebrand Factor (vWF; rabbit polyclonal antibody; dilution, 1:100; DakoCytomation, Glostrup, Denmark). The basement membrane was identified with an antibody against mouse type IV collagen (rabbit polyclonal 1:1000; Cosmo Bio, Tokyo, Japan). Pericytes were identified with antibodies to α-smooth muscle actin (Cy3-conjugated mouse monoclonal, 1:500; Sigma-Aldrich) and desmin (rabbit polyclonal, 1:200, Abcam, Cambridge, MA, USA). VASH2-expressing cells were labeled with a rabbit polyclonal antibody to mouse VASH2[
19] (1:100; a generous gift from the Institute of Development, Aging and Cancer, Tohoku University, Sendai, Japan). After several washes with PBS, specimens were incubated with combinations of fluorescent (FITC, Cy3, and Cy5)-conjugated anti-rat, -hamster, and -rabbit secondary antibodies (Jackson ImmunoResearch, West Grove, PA, USA) for 2 h at room temperature. Immunostained sections were examined using a Leica TCS-SL confocal laser-scanning microscope (Leica Microsystems, Wetzlar, Germany).
Immunolabeling of semi-thin sections
After thawing and air-drying, cryosections were rehydrated in PBS then incubated with 4% Block Ace blocking solution (Dainippon Seiyaku, Tokyo, Japan) to reduce nonspecific background staining. For ordinary immunoenzymatic staining, tissue sections were incubated with anti-CD31 overnight at 4°C. The sections were further incubated with goat anti-rat immunoglobulins labeled with horseradish peroxidase (HRP) (GE Healthcare UK, Buckinghamshire, UK; 1:100 in PBS with 1% heat-inactivated normal mouse serum) for 2 h. The HRP reaction was developed at RT for 20 min in a solution of 10 mg of 3′-diaminobenzidine hydrochloride (DAB: Dojin Chemicals, Kumamoto, Japan) in 30 ml of PBS with 10 μg of 30% H2O2. Sections were washed in distilled water. After the DAB reaction, the sections were fixed in 2.5% glutaraldehyde in PB at 4°C for 1 h and subsequently in 2% osmium tetroxide in PB at room temperature for 1 h. They were then dehydrated in a graded series of ethanol and embedded in an epoxy resin. Semi-thin sections stained with 0.05% toluidine blue were examined using a light microscope.
Morphometric analysis
For analyses of the microvessel density (MVD), the total areas of CD31, vWF-positive capillaries, and venules were assessed by scanning tumor sections under × 40 magnification and counting in 10 random fields under × 600 magnification[
44,
45]. Pericytes were identified by scanning tumor sections under × 1000 magnification and counting in 3 random fields under × 1000 magnification[
46]. These data were analyzed using a BZ-Analyzer (Keyence, Osaka, Japan).
Human tissue samples
Surgical specimens were obtained from 10 patients with colorectal adenoma (n = 5) or adenocarcinoma (n = 5) who underwent proctocolectomies in the Department of Surgery at Nishiarai Hospital (Tokyo, Japan) between April 2010 and March 2011. Patients with additional cancers were excluded. The Clinical Pathology Department of the Nishiarai hospital confirmed the histopathological diagnosis. Written informed consent was obtained from all patients for the surgery and for the use of their resected samples. H&E staining was performed to determine the histologic tumor type, lymphatic invasion, and vascular invasion in all specimens.
Western blot analysis
Tissue and cell samples were lysed in SDS sample buffer, separated in 10% SDS-acrylamide gels, and electrotransferred to nitrocellulose membranes. After blocking with 5% non-fat dry milk in TBST buffer (10 mmol/L Tris–HCl (pH 8.0), 150 mmol/L NaCl, 0.05% Tween 20), the nitrocellulose membranes were probed with anti-CD31 (1:200; Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-VEGF (1:1,000; R&D Systems, MN, USA), anti-VEGF (1:1,000; R&D Systems, MN, USA), anti-CEA (1:2,000; Santa Cruz Biotechnology), anti-CA19-9 (1:2,000; Santa Cruz Biotechnology), anti-KRAS (1:2,000; Abcam, Cambridge, MA, USA), anti-p53 (1:1,000; R&D Systems), and anti-α-tubulin (1:1,000; Santa Cruz Biotechnology) antibodies, followed by incubation with HRP-conjugated anti-rabbit or anti-rat immunoglobulin G secondary antibodies (1:2,000; Jackson ImmunoResearch). The antibody binding was then visualized with enhanced chemiluminescence reagents (GE Healthcare, Amersham, UK), and the band images detected using the LAS3000 system (Fuji Film, Tokyo, Japan) were densitometrically analyzed using Image Gauge (Fuji Film).
Quantitative real-time PCR
Total RNA was extracted using QIAzol Lysis Reagent (Qiagen, Venlo, Netherlands). First-strand cDNA was generated using ReverTra Ace (Toyobo, Osaka, Japan). Quantitative real-time RT-PCR was performed using the CFX96 real-time PCR detection system (Bio-Rad Laboratories, Hercules, CA, USA) according to the manufacturer’s instructions. PCR conditions consisted of an initial denaturation step at 95°C for 3 min, followed by 40 cycles of 10 s at 95°C, 10 s at 56°C, and 30 s at 72°C. Relative mRNA levels of target genes were normalized to the beta-2-microglobulin (B2m) mRNA level. The primer pairs used were as follows: mouse B2m forward, 5′-GGTCTTTCTGGTGCTTGTCTCA-3′, and reverse, 5′-GTTCGGCTTCCCATTCTCC-3′; mouse CD31 forward, 5′-TTCAGCGAGATCCTGAGGGTC-3′, and reverse, 5′-CGCTTGGGTGTCATTCACGAC-3′; mouse CD105 forward, 5′-TACAGTGCATCGACATGGAC-3′, and reverse, 5′- TCAGAGGTCAATGGAGACAC-3′; mouse Vegfa forward, 5′-AGAGAGCAACATCACCATGC-3′, and reverse, 5′- TCTGAACAAGGCTCACAGTG-3′; mouse Vash2 forward, 5′-GGACATGCGGATGAAGATCT-3′, and reverse, 5′- CTAGATCCGGATCTGATAGC-3′.
X-gal staining
Frozen sections were incubated in the dark for 18 h at 37°C in X-gal solution containing 1 mg/mL 5-bromo-4chloro-3-indolyl-β-d-galactoside (X-Gal, Gene Therapy Research Reagents, San Diego, CA, USA).
Transmission electron microscopy
Anesthetized mice were fixed by vascular perfusion of 4% PFA and 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (100 mL; pH 7.4) at a pressure of 120 mmHg. Immediately after the perfusion, the tumor tissues were removed, cut into small pieces, and immersed in the same fixative for another 2 h at 4°C. Specimens were then treated with 1% OSO4 for 2 h at 4°C, and then with saturated uranyl acetate for 3 h at room temperature. Thereafter, specimens were dehydrated in a graded series of ethanol and embedded in epoxy resin. Ultrathin sections (70 nm thick) were made, counterstained with saturated uranyl acetate followed by lead citrate, and observed using a Hitachi H-7000 electron microscope (Hitachi High-Technologies Co., Tokyo, Japan).
Statistics
All results are expressed as mean ± standard deviation (SD). The statistical significance of differences was determined using the one-tailed Student’s t-test. The difference between two values was considered statistically significant if the P value was less than 0.05, and as highly significant if the P value was less than 0.01.
Acknowledgments
We thank Dr. P. Baluk (UCSF, San Francisco, USA) for comments on the manuscript and Ms. K. Nakada, Mrs. H. Sagawa, Mrs. Y. Yamazaki, and Mrs. K. Motomaru of Tokyo Women’s Medical University for their technical help. We are grateful to Dr. T. Shimakawa, Director of Surgery at Nishiarai Hospital, for the generous donation of pathological tissues. Special thanks go to Prof. M. Fukumoto, Dr. K. Kuwahara, and Dr. M. Fukumoto of the Department of Pathology at Institute of Development, Aging and Cancer, Tohoku University, Sendai, Japan for their advice and expertise.
Funding
This work was mainly supported by a Grant-in-Aid for Young Scientists (Start-up) (No.20890231) from the Japan Society for the Promotion of Science, and by a Grant-in-Aid for Scientific Research (B) (No.19390053) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, and the Global COE program, the Multidisciplinary Education and Research Center for Regenerative Medicine (MERCREM), from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) Japan, and Technology of Japan, and by the Cooperative Research Project Program of Joint Usage/Research Center at the Institute of Development, Aging and Cancer, Tohoku University.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
SK and YS participated in the study design, performing the experiments, data analysis and drafting of the manuscript. KS and SM assisted with the in vitro, in vivo and cloning experiments and provided technical assistance. MM and SK provided material. AY performed immunohistochemistry experiments. YS was involved in the conception of the study and drafting the manuscript. TE participated in the study design, data analysis and writing of the manuscript. All authors read and approved the manuscript.